GPU Supply Agreements
Definition
GPU supply agreements are contracts governing the purchase, allocation, delivery, financing, or hosting of AI accelerators used in training and inference infrastructure.
Why it matters
AI infrastructure economics can depend on whether a provider actually controls the GPUs it markets. Delivery timing, title, warranties, export restrictions, and financing liens all affect investor risk.
Common misconceptions
- •An allocation, purchase order, reservation, or reseller quotation is not necessarily a binding commitment to deliver specific hardware on a fixed date.
- •Prepayment does not guarantee title or priority if the seller fails, inventory is not segregated, or another lender has a competing lien.
- •Equivalent substitution can change memory, networking, cooling, software, or customer economics even when nominal compute performance appears similar.
- •Hardware delivery alone does not create revenue; power, racks, networking, cooling, software, installation, and customer acceptance remain critical-path dependencies.
Technical details
Terms to inspect
Confirm buyer, seller, model type, quantity, delivery schedule, title transfer, warranty coverage, cancellation rights, and substitution rights.
Check whether the GPUs are owned, leased, financed, pledged as collateral, or merely reserved through a reseller.
Look for restrictions on resale, export, redeployment, and use by sanctioned or sensitive end customers.
Investment relevance
A provider with signed customer contracts but uncertain GPU delivery has different risk than a provider with installed hardware, power, network, cooling, and customer utilization already in place.
Delivery and acceptance
Map deposits, manufacturing allocation, shipment, customs, installation, testing, customer acceptance, and final payment. Define delay remedies, liquidated damages, cancellation, force majeure, partial delivery, and long-stop dates. Revenue and debt draws should follow usable installed capacity rather than an announcement or expected ship date.
Title, liens, and counterparty risk
Determine when title and risk of loss transfer, whether units are identified by serial number, where deposits sit, and whether the buyer receives a purchase-money or other perfected interest. Review manufacturer, distributor, reseller, lessor, and financing relationships because the contracting seller may not control inventory or warranty performance.
Technology and regulatory change
Review substitution rights, price resets, end-of-life notices, warranty transfer, export licensing, geographic restrictions, sanctioned end users, and permitted resale. A delayed contract can become uneconomic if a newer generation reaches market before delivery, while restrictive terms may prevent redeployment to the highest-value customer or jurisdiction.
Capacity-to-revenue bridge
For GPU supply agreements, bridge physical capacity to billable revenue. Start with contracted or announced units, then deduct capacity not yet delivered, powered, cooled, networked, commissioned, accepted by customers, or available after redundancy and maintenance requirements.
Build a monthly schedule for installed capacity, usable capacity, committed capacity, billed capacity, and collected revenue. This prevents double-counting the same GPU, rack, or megawatt across marketing pipeline, financing collateral, and customer backlog.
Separate high-margin infrastructure revenue from pass-through power, setup fees, burst usage, credits, taxes, and reimbursed costs. Revenue quality depends on margin, duration, collectability, and renewal probability, not only gross contract value.
Contract and counterparty diligence
Review the exact contracting party, guarantor, minimum commitment, ramp schedule, delivery conditions, service levels, termination rights, cure periods, force majeure, assignment rights, deposits, and lender step-in rights.
Customer quality matters because AI demand can be volatile. Underwrite concentration, funding runway, payment history, use case, workload portability, and whether the customer can switch to hyperscalers or newer hardware.
Supplier diligence should cover title transfer, liens, serial-number evidence, warranty, replacement rights, export controls, delivery delay remedies, and whether a reseller actually controls the inventory it promises.
Operating constraints and cost stack
AI compute economics are constrained by power price, power availability, cooling design, rack density, network fabric, facility uptime, maintenance, software orchestration, spare parts, and labor. A GPU fleet can be technically installed but commercially weak if one of these constraints binds.
Stress power-price increases, curtailment, delayed interconnection, transformer lead times, cooling retrofits, customer credits, lower utilization, and hardware failures. Compare gross utilization with contribution margin after power and operating costs.
For financing, match customer contract tenor and hardware useful life to debt amortization. A long loan against short-lived or rapidly repricing hardware can leave residual-value risk with the lender or vehicle.
Refresh, residual value, and monitoring
Track hardware by cohort: model, purchase date, installed cost, memory profile, networking, warranty, utilization, average realized rate, power draw, and expected resale or redeployment value.
Monitor competitive GPU pricing, new chip launches, customer workload shifts, inference versus training mix, cloud spot pricing, and resale market depth. A unit that still functions can become economically stale before physical failure.
Warning signs include revenue booked before acceptance, unclear ownership of hardware, repeated delivery delays, rising service credits, power constraints, low realized utilization, customer nonpayment, and capex needs that are not reflected in the financing model.